Our empirical results indicate that the static pooled probit model is strongly dominated by the alternative models with serial dependence. However, state- dependence and transitory country-specific errors are essentially observationally equivalent. Only if we include random time-specific effects into the model with state-dependence, we find that both sources of serial dependence are significant, even though the time-specific effect is small with limited effect on the overall fit of the model. On the other hand, our assessment of the ability to predict currentaccountreversals provides strong support for the model with transitory country- specific errors and without state-dependence, which appears to present the best compromise between log-likelihood fit and predictive performance. Also, we do not find conclusive evidence for the existence of random country-specific effects. Overall, our results relative to the determinants of currentaccountreversals are in line with the those in the empirical literature on currentaccount crises and confirm the empirical relevance of theoretical solvency and sustainability considerations w.r.t. a country’s trade balance. In particular, countries with high currentaccount imbalances, low foreign reserves, a small fraction of concessional debt, and unfavorable terms of trades are more likely to experience a currentaccount reversal. These results are fairly robust against the dynamic specification of the model.
An analysis of the Malawi’s currentaccount balance indicates that it has been in persistent deficit largely dominated by the merchandise trade balance. The study used the cointegration analysis to identify the long-run determinants of Malawi’s currentaccount deficit and the error correction model to examine the short-run dynamics. The study also used impulse response analysis to capture dynamic interactions among the variables. Generally the results indicate that external factors which are openness, terms of trade, accumulation of external debt and the liberalization of currentaccount, fundamentally determined the currentaccount deficit in Malawi during the period under study. Impulse response analysis indicate that the exogenous shocks to the real exchange rate persistently worsens the currentaccount deficit and forecast error variances demonstrates that even at three years real exchange rate can explain significant proportion of the currentaccount, and the explanatory strength is increasing with time. These findings suggest that government can directly control the behaviour of currentaccount through exchange rate policy hence underscoring the importance of exchange rate policy in managing the currentaccount deficits.
Khan and Knight (1983) carried out an empiricalanalysis on determinants of currentaccount balances of non-oil developing countries in the 1970s. The study estimated a simple currentaccount model whereby currentaccount balance (excluding official transfers), as a ratio of nominal exports of goods, is a function of terms of trade, growth of real gross national product in industrial countries, foreign real interest rate, real effective exchange rate, fiscal position (as a ratio of nominal gross domestic product) and linear time trend. The equation was estimated using pooled time-series cross-section data for the sample of 32 non-oil developing countries. It was revealed that both external and internal factors affect currentaccount balance of non-oil developing countries, whereby terms of trade, growth of industrial countries and fiscal balance have positive effects, while foreign real interest rate and real effective exchange rate negatively affect currentaccount balance. Based on the coefficients, it was evident that the most important explanatory variable is the terms of trade, while foreign real interest rate, the real effective exchange rate, and the government’s fiscal position turn out to be of roughly equally importance. The least important factors are growth in the industrial countries and the time trend as determinants of currentaccount developments in this particular sample of non-oil developing countries (Khan and Knight, 1983).
Barro (2001) analyses the effect of currency and banking crises on growth and investment in 9 East Asian countries using a five-year grouped panel from 1980 to 2000 for 67 countries. The Author uses three stage least square without country fixed effects as the estimation procedure and investment, initial GDP, male upper-level schooling, life expectancy, a total fertility rate, government consumption, a rule-of-law index, openness, inflation and a growth rate of terms of trade as control variables. He finds that a combined currency and banking crisis typically reduces economic growth over a five-year period by 2 % per year, compared with 3 % per year for the 1997-98 crisis in East Asia. Further, he explores dynamics using lagged dummies. The broader analysis found no evidence that financial crises had effects on growth that persisted beyond a five-year period. However, when analyzing the effect of banking and currency crises separately the estimates suggest that both may have small positive effect on growth. Additionally, estimates of the lagged banking crisis’ dummy in investment equation suggest small negative effect on investment.
Chinn and Prasad (2003) emphasize that instead of capital controls country’s financial development is positively correlated with currentaccount balance in developing coun- tries. Contradicting this result, Cheung et al. (2013) and Mendoza et al. (2009) show that financial sector development measured by private credit ratio has a negative impact on currentaccount balance. Such negative relation might arise from the fact that developed financial system and legal investment protection regulation may divert the capital flows into other countries with more liquid assets and competitive market (Bernanke 2005; Ju and Wei 2006). Similarly, Alfaro et al. (2008) confirm that domestic and international market imperfection, low institutional quality and weak governance structure in devel- oping countries increase the investor’s high risk of return. Thus, capital is uphilling in the relatively more stable and developed financial system particularly in European and North American economies (Caballero et al. 2008). Besides, based on an empirical research of a panel of developing countries Calderon et al. (2002) reveal that high cur- rent account deficit tends to associate with output growth, in terms of trade shock and currency appreciation. However, past global economic shocks such as Asian crisis, Latin American crisis and recent financial crisis reduce the investment levels (Reinhart and Rogoff 2008; Chinn and Ito 2007; Eichengreen 2006).
In this paper I analyze the anatomy of currentaccount adjustments in the world economy during the last three decades. The main findings may be summarized as follows: (a) Major reversals in currentaccount deficits have tended to be associated to “sudden stops” of capital inflows. (b) The probability of a country experiencing a reversal is captured by a small number of variables that include the (lagged) currentaccount to GDP ratio, the external debt to GDP ratio, the level of international reserves, domestic credit creation, and debt services. (c) Currentaccountreversals have had a negative effect on real growth that goes beyond their direct effect on investments. (d) There is persuasive evidence indicating that the negative effect of currentaccountreversals on growth will depend on the country’s degree of openness. More open countries will suffer less n in terms of lower growth n than countries with a lower degree of openness. (e) I was unable to find evidence supporting the hypothesis that countries with a higher degree of dollarization are more severely affected by currentaccountreversals than countries with a lower degree of dollarization. And, (f) the empiricalanalysis suggests that countries with more flexible exchange rate regimes are able to accommodate the shocks stemming from a reversal better than countries with more rigid exchange rate regime.
In parallel, the paper introduces in the discussion of the currentaccount the policy- related question of external sustainability, particularly relevant for the period since 1999. We define external sustainability, independently of our empirical model, as the currentaccount to GDP ratio that stabilises the net foreign asset position (or alternatively, external debt for simplicity) to GDP ratio. A novelty of the paper, in this respect, is that it relates the policy measures needed to restore external sustainability to the specific variables of the equilibrium currentaccount model that is developed. A further contribution of the paper is the use of a variety of econometric tests for both the long-run analysis and the short-run dynamics in order to account for shifts in behaviour during times of significant structural change. This is particularly true in the case of Greece, which in the period under review went through a process of financial liberalisation and policy regime changes that substantially altered the country’s macroeconomic conditions. The paper concludes by exploring possible policy options for reverting to external sustainability, while illustrating with an example of comparative static analysis how the estimated equilibrium model could be applied to simulate different adjustment paths.
The purpose of this paper is twofold. First, it attempts to empirically determine the main variables that influence the currentaccount in Greece both in the long run and the short run using co-integration analysis. In this respect we find that a stable equilibrium currentaccount model can be derived if the ratio of private sector financing to GDP, as a proxy for financial liberalization, is included in the long-run specification. Secondly, on the basis of the empirical results, it addresses the sustainability question, particularly for the period since 1999. A novelty of the paper is that it relates the policy measures needed to restore sustainability to the equilibrium model of the currentaccount that is developed. A further contribution of the paper is the use of a variety of econometric tests for the long-run analysis and the short-run dynamics in order to account for shifts in behaviour during times of significant structural change. This is particularly true in the case of Greece, which in the period under review went through a process of financial liberalization and policy regime changes that substantially altered the country’s macroeconomic conditions. The paper concludes by exploring possible policy options for reverting to external sustainability, using the estimated equilibrium model to perform a simulation exercise.
It seems that the key motivation to shut down the currentaccount channel as a dynamic shock- propagation mechanism is to keep the analysis simple. This is accomplished by either incorporating the complete asset markets assumption 2 (e.g., Clarida, Gali and Gertler (2001), Gali and Monacelli (2002), Chari, Kehoe and McGrattan (1998) and Devereux and Engel (2000)) or by imposing a unitary elasticity of substitution between domestic and foreign goods 3 (e.g., Corsetti and Pesenti (2001a, 2001b), Tille (2001)). However, as pointed out by Obstfeld and Rogoff (1995b), the assumption of complete asset markets is not realistic in a model with imperfections and rigidities in goods market because with nominal rigidities monetary policy will affect real variables including the currentaccount. Thus, with incomplete asset markets the dynamics of currentaccount do matter for monetary policy because then, besides dealing with the distortions created by monopolistic competition, the central bank need to address the inefficiencies caused by incomplete asset markets. Moreover, in an empirical paper, Lane and Milesi- Ferretti (2002) link net foreign asset positions to long-run values of real exchange rates and suggest that optimal monetary policy responses may depend on the movements in the currentaccount.
On the other hand, another line of research supports the validity of the Feldstein and Horioka’s methodology in measuring capital mobility, and they explain the puzzle on methodological and econometric grounds. Within this framework, a number of researchers focus on the role of policy regime changes. (Gundlach and Sinn 1992, Jansen 1996, Jansen and Schulze 1996, Sarno and Taylor 1998, Bajo- Rubio 1998, Ozmen and Parmaksiz, 2003, 2005, and Coakley et al. 2004). Their findings suggest that policy regime changes introduce structural breaks which significantly bias the empirical results towards rejecting the hypothesis of capital mobility. Such evidence calls for a “country by country” approach –as opposed to cross section analysis- in order to ensure that the particular characteristics of the economy under examination are incorporated explicitly into the empiricalanalysis (Corbin 2001, Coakley et al. 2004, Taylor 2002, Jansen 1996, Mark 2003, Giannone and Lenza 2004, provide an analysis of the effects of country heterogeneity on the estimation methodology). We follow this line of research and investigate the role of policy regime changes in applying the Feldstein and Horioka methodology in the case of the Greek economy.
He uses three stage least squares without country fixed effects as the estimation procedure and employ investment, initial GDP, male upper-level schooling, life expectancy, a total fertility rate, government consumption, a rule-of-law index, openness, inflation and a growth rate of terms of trade as control variables. He finds that a combined currency and banking crisis typically reduces economic growth over a five-year period by 2 % per year, compared with 3 % per year for the 1997-98 crisis in East Asia. Further, he explores dynamics using lagged dummies. The broader analysis found no evidence that financial crises had effects on growth that persisted beyond a five-year period. However, when analyzing the effect of banking and currency crises separately the estimates suggest that both may have small positive effect on growth. Additionally, estimates of the lagged banking crisis’ dummy in investment equation suggest small negative effect on investment.
In this paper I have shown that sudden stops and currentaccountreversals have been closely related. The econometric analysis suggests that restricting capital mobility does not reduce the probability of experiencing a reversal. Currentaccountreversals, in turn, have had a negative effect on real growth that goes beyond their direct effect on investment. The regression analysis indicates that the negative effects of currentaccountreversals on growth will depend on the country’s degree of trade openness: More open countries will suffer less – in terms of lower growth relative to trend– than countries with a lower degree of trade openness. On the other hand, the degree of financial openness does not appear to be related to the intensity with which reversals affect real economic performance. The empiricalanalysis also suggests that countries with more flexible exchange rate regimes are able to accommodate better shocks stemming from a reversal than countries with more rigid exchange rate regimes. In interpreting the findings
In undertaking this empirical exercise, we attempt to characterize a broad set of stylized facts associated with reversals and crises. However, caution must be exercised in interpreting these regularities as a reliable predictive model. The burgeoning analytical literature of financial crises has highlighted several mechanisms that can generate such an outcome: inconsistency between deteriorating fundamentals and the maintenance of a fixed exchange rate (Krugman (1979)), self- fulfilling crises à la Obstfeld (1994), models of crises based on bank runs à la Diamond and Dybvig (1983) (Goldfajn and Valdés (1997), Chang and Velasco (1998)). Although these mechanisms generating crises are different, the models point to an overlapping set of indicators (e.g. the level of reserves, the rate of growth in domestic credit, world interest rates etc). Hence empirical exercises relating the probability of a crisis to a large set of indicators cannot discriminate between different explanations for crises. Failure to identify the (potentially different) mechanisms underlying crises limits the usefulness of these exercises as predictive tools because the reduced-form relationship between crisis events and indicators averages the particular pattern of crises prevailing in the sample, which may not be repeated in the future (as in the standard Lucas critique). In addition, policy inference is hindered by the fact that the crisis-generating mechanisms, which we cannot disentangle, can have different policy implications (eg, tight monetary policy is called upon in a standard Krugman-type crisis, while a more flexible monetary policy is called upon in the event of bank runs). 2. THEORETICAL DETERMINANTS OF REVERSALS AND CURRENCY CRISES
The motivation for this paper is the uncovering of some new facts. Figures 2 to 5 propose a simple decomposition of the trade balance into the goods balance and service balance components for the same four European countries included in Figure 1. While for Germany (Figure 2) the large trade surplus emerges from a trade surplus in goods, accompanied by a trade deficit in services, the opposite is true for the other countries. Spain, Portugal, and Greece, in fact, exhibit increasing trade deficits in goods, but surpluses in services (Figures 3, 4, and 5). 1 Moreover, also looking at the bilateral trade relationships between Germany and Spain, Portugal, and Greece we observe the same pattern: these southern European countries display trade surpluses in services and trade deficits in goods (Figures 6, 7 and 8). Starting from this motivating evidence, the contribution of this paper is to propose a theory of how asymmetric trade liberalization processes can affect currentaccount dynam- ics. I start by outlining a simple theoretical model where I show how asymmetric trade liberalizations can affect currentaccountdynamics. I then propose a quantitative analysis of the German surplus using a standard 2 country international real business cycle model with trade costs, and I show how the asymmetry in the liberalization of manufacturing versus ser- vice trade documented in Barattieri (2014), can explain a significant fraction of it. Finally, I propose some empirical evidence that broadly support the main predictions of the theory both using data from 24 OECD countries and the BRICS and using a sample of developing
The bene…ts of treating the savings rate as an exogenous constant are twofold. Firstly, we are able to provide a clear exposition of the intertem- poral e¤ects of tax reform on the growth rate of the economy and real variables, such as the currentaccount balance (see Figure 1). The long run e¤ects of tax reform are con…ned to level e¤ects, consistent with re- cent empirical evidence. Secondly, if we were to endogenize the savings rate, Turnovsky’s (1996) assumption of an in…nite-lived agent is restrictive to any welfare analysis of tax reform since it implies welfare maximisation
variables on external imbalances. Leigh (2008) finds that a increase in government consumption is related with an appreciation of the equilibrium real exchange in case of both developing and advanced economies by using panel estimation. The actual impact on the currentaccount could vary depending on the dynamic adjustment path of the actual real exchange rate toward the equilibrium; large currentaccount worsening can obtain if the real exchange rate appreciates above its equilibrium level that is overshooting. Mohammadi (2004) finds broadly symmetrical impact for fiscal expansions and contractions for a sample of 20 advanced and 43 emerging and developing economies that a tax-financed spending increase is associated with a currentaccount worsening both for developing and developed countries and the currentaccount balance worsens more if the spending is bond-financed in case of developing economies rather than developed ones. The study done by Khalid and Guan (1999) findings does not support any long-run relationship between the currentaccount deficit and the fiscal deficit for advanced economies, while the data for developing countries does not reject such a relationship. However, their results suggest a causal relationship between the fiscal and currentaccount balances for most countries in their sample, running from the budget balance toward the currentaccount balance.
Another important result was the finding that the estimates of both the parameter and the degree of habit formation showed little sensitivity to interest rate changes, as already shown in other works in the literature on currentaccountdynamics. In estimating the degree of taxation of the economy, we found a value of around 10%, although a significant linear correlation between the change in government expenditure and output variation was not identified. The series which was estimated from the expression incorporating the rule of thumb consumption and habit formation concepts was compared to the actual series and the optimal currentaccount series. The latter was obtained from the strategy used by Campbell and Shiller (1987) to estimate the optimal behavior of the currentaccount. The comparison indicated that the curve estimated with the inclusion of rule of thumb consumer is consistent with the other two, though there is departure of the estimated curve from the other two in some years of the 1970s and 1990s. The curve presented an estimated standard deviation larger than that of the actual curve, suggesting higher volatility. Furthermore, the correlation between the estimated curve and actual curve was lower than the value obtained when the optimal and actual series were taken into consideration.
Debelle and Faruqee, (1996) investigate the factors affecting currentaccount balances by using data of 21 industrial countries over the period of 1971 to 1993.They use panel data regression technique & error correction model. Results show that capital controls, terms of trade and fiscal surplus do not play a significant role in the variation of a currentaccount in long term, while government debt, relative income and demographic have significant effect on currentaccount balance. They also conclude that the changes in fiscal policy, movements in terms of trade, state of business cycle and the real exchange rate are the factors which having the significant impact on currentaccount balance in short term.
This ﬁnancial friction enables us to check whether the sensitivity of the currentaccount with respect to the degree of competition is dependent on the non-stationarity of the currentaccount. The simulation results given below demonstrate clearly that the currentaccount response to a monetary shock is sensitive to the degree of competition, irrespective of whether there is or not a unit root in the foreign asset position. The intuition behind this result is that the long-run asset position can, at most, aﬀect current expenditure levels (via intertemporal substitution and consumption smoothing), but does not aﬀect the current expenditure switching. It is the latter and not the former that determines the sign of the currentaccount response in our symmetric model. This presumption is conﬁrmed by sensitivity analysis (not shown), where the degree of intertemporal substitution as well as the steady-state asset return are increased one hundred fold. Given these parameter values, there was no change in the qualitative response of the currentaccount.